[Python] How to show error output between boosting rounds in XGBoost?

I'm using Scikit-Learn API, XGBRegressor. I'm trying to make my model as verbose as possible. These are the parameters of the model. This is running on Kaggle kernels. df_train and df_target are pandas dataframes.

model = xgb.XGBRegressor(
n_estimators=2**8,
max_depth=5,
learning_rate=0.04,
subsample=0.9,
colsample_bytree=0.9,
objective='reg:squarederror',
booster='gbtree',
importance_type='weight',
tree_method='gpu_hist',
silent=False,
random_state=SEED
)


and this is the parameters of fit(). I have to see training rmse between boosting rounds like in the LightGBM. Does XGBoost have that functionality?

model.fit(df_train, df_target, eval_metric='rmse', eval_set=[(df_train, df_target)], verbose=True)

Continue reading...